Body cameras were on. The footage was released, on average, 71 days later.

Body-worn cameras are now near-universal in large departments. What happens to the recording afterward varies enormously, and written policy turns out to be a poor predictor of it.

MNC News reviewed 210 critical incidents over five years across 63 agencies — cases involving death or serious injury, where nearly every department’s own policy requires release. The median time from incident to public release was 71 days. The median policy deadline in those same agencies was 52.

The only thing that predicted speed

Departments were sorted by every variable available: size, budget, whether the policy was written by the agency or by a civilian board, whether footage went to a prosecutor first. Almost none of it explained the spread.

One thing did. Agencies in states with a statutory release deadline enforced by a court — rather than an internal policy deadline — released footage in an average of 19 days. Agencies whose deadline lived only in their own policy manual averaged 84.

A deadline you enforce on yourself is a goal. A deadline someone else enforces is a deadline. Departments know the difference, and so does everyone waiting.— A former internal affairs commander who reviewed the findings

The cases with no footage

In 38 of the 210 incidents, no footage was released at all. In 22 of those, the stated reason was an open investigation that remained open at the time of our review — in nine cases for more than three years. In eleven, the department said the camera was not recording. In five, the recording existed but was withheld under a privacy exemption covering people who were not the subject of the incident.

• Reviewed: 210 critical incidents across 63 agencies over five years
• Median release: 71 days; median internal policy deadline: 52 days
• With a statutory, court-enforced deadline: 19 days on average
• No footage released: 38 incidents, including 22 citing an open investigation

Several chiefs pushed back on the framing, noting that early release can compromise a prosecution and that a redaction pass on a two-hour recording is genuinely slow work. Both points are fair, and neither explains why the same work takes four times as long in one state as in the one next to it.